The introduction of density functional theory(DFT)and electronic structure has brought computational methods into the field of materials science.In these theoretical calculations,quantum mechanics is predominantly use...The introduction of density functional theory(DFT)and electronic structure has brought computational methods into the field of materials science.In these theoretical calculations,quantum mechanics is predominantly used.Machine learning(ML)and high-throughput computing share some inherent similarities,as both can extract valuable information from massive datasets and possess parallelism and scalability.ML techniques simulate human thought processes,with algorithms that make decisions and have good scalability and strong generalization abilities.The combination of high-throughput and ML technologies leverages the advantages of high-throughput technology standardization and high capacity,addressing the challenges faced by ML at the front end.This complementary combination is expected to further enhance the efficiency of material screening and development.In data mining,using ML methods on various databases,the interrelationships between molecular structures and properties are discovered from large amounts of data.Mapping,current utilization of DFT,materials genomics,and high-throughput computing have generated a substantial amount of data.This review provides new insights into the development of electrochemistry.展开更多
基金supported by the National Natural Science Foundation of China(grant no.U1904215)the Natural Science Foundation of Jiangsu Province,China(grant no.BK20200044)the Changjiang Scholars Program of the Ministry of Education,China(grant no.Q2018270).
文摘The introduction of density functional theory(DFT)and electronic structure has brought computational methods into the field of materials science.In these theoretical calculations,quantum mechanics is predominantly used.Machine learning(ML)and high-throughput computing share some inherent similarities,as both can extract valuable information from massive datasets and possess parallelism and scalability.ML techniques simulate human thought processes,with algorithms that make decisions and have good scalability and strong generalization abilities.The combination of high-throughput and ML technologies leverages the advantages of high-throughput technology standardization and high capacity,addressing the challenges faced by ML at the front end.This complementary combination is expected to further enhance the efficiency of material screening and development.In data mining,using ML methods on various databases,the interrelationships between molecular structures and properties are discovered from large amounts of data.Mapping,current utilization of DFT,materials genomics,and high-throughput computing have generated a substantial amount of data.This review provides new insights into the development of electrochemistry.
